This looks like a really cool work for a beginner and I do not mean to diminish his effort in any way.
But face recognition is a sensitive/politically-charged topic, I know several grad students (including me) who inspite of having ready-to-deploy scalable software and datasets (~10M instagram images) stay away from doing this type of a demo because reputational risks are enormous. Consider the controversy around geofeedia etc. The last thing you want as a PhD student is press interpreting your research incorrectly and blaming you for causing widespread harm. It happened to a student/professor in my department and even then the infamous study in question was in collaboration with the social network.
There are several cool alternatives, apply it to movies, video game faces etc.
Although they often have less data (but it wouldn't really matter in this case since he used a pretrained model), he could have used one of the many public face recognition datasets that are already out there and exist exactly for this purpose like LFW [0].
You probably shouldn't be posting images with captions like "random ig girl."
Via the Instagram Terms of Use:
> We prohibit crawling, scraping, caching or otherwise accessing any content on the Service via automated means, including but not limited to, user profiles and photos (except as may be the result of standard search engine protocols or technologies used by a search engine with Instagram's express consent).
Hasn't OP already addressed this in the 1st few lines in his post ?
> But after showing it to a couple of my "friends" they thought it was too creepy and Instagram might sue me for breaking their platform policy and I should stop doing it.
> So, I did what most sane people would do - write a blog post detailing how I did it, and open source it.
Or did he update his blog post after reading the comments here? I don't even know if there's any way to find out.
Side note: I don't find this creepy. It's somewhat ironic that Facebook, the owner of Instagram started off way more creepy when Zuckerberg created Facemash, the precursor to Facebook.
> For those who haven’t seen the movie, Facemash presented the user with two randomly selected pictures of Harvard students and then let the user vote on which one was hotter.
Is being globally public enough reason to think its not creepy? I'm pretty sure the road outside my house is public, but it is still creepy if someone in a clown costume stands motionless in it and points at my bedroom window at 3am every morning.
But face recognition is a sensitive/politically-charged topic, I know several grad students (including me) who inspite of having ready-to-deploy scalable software and datasets (~10M instagram images) stay away from doing this type of a demo because reputational risks are enormous. Consider the controversy around geofeedia etc. The last thing you want as a PhD student is press interpreting your research incorrectly and blaming you for causing widespread harm. It happened to a student/professor in my department and even then the infamous study in question was in collaboration with the social network.
There are several cool alternatives, apply it to movies, video game faces etc.